- Top of page
- PATIENTS AND METHODS
For many years enzymatic serum acid phosphatase (AP) was the only prostate tumour marker used by urologists in detecting and evaluating a patient with prostate cancer . Then PSA was discovered and found to be more sensitive than AP, especially than the newer form generally referred to as prostate AP (PAP, detected by radioimmunoassay)  with the logical result that PSA has become the most commonly used prostate cancer marker.
However, some studies [3,4] suggested that serum levels of enzymatic serum AP are directly related to pathological stage, and more recently others have shown that PAP  and enzymatic serum AP  can predict biochemical recurrence after radical retropubic prostatectomy (RRP). In contrast, others have suggested that PAP is no longer needed when evaluating a patient diagnosed with prostate cancer, because it has no role as a marker of prostate cancer staging  and because PAP is less sensitive than PSA in predicting biochemical recurrence after RRP or radiotherapy .
On the other hand, alkaline phosphatase (ALP) is well known as a marker of bone and hepatobiliary disease. Its bone isoenzyme (BAP) is abnormally elevated and correlated with the number of bone metastases in prostate cancer  secondary to increased osteoblastic activity . However, the levels of ALP are in the normal range in the presence of localized prostate cancer.
The present long-term retrospective study was designed to redefine the role of the enzymatic serum total AP (TAP), its tartrate-inhibitable prostatic fraction (PFAP) and ALP as markers of the severity of prostate cancer, and their role as predictors of biochemical failure after RRP for localized disease in the present era of PSA testing.
PATIENTS AND METHODS
- Top of page
- PATIENTS AND METHODS
We retrospectively reviewed the clinical data and pathology reports of 195 consecutive patients with clinically localized prostate cancer who underwent anatomical RRP with standard obturator lymph-node dissection from 1 January 1990 to 31 December 1995, by one urologist (W.C.D.) at Beth Israel Deaconess Medical Center. Patients who had a history of neo- or adjuvant hormonal therapy were excluded from the study.
Prostatic needle biopsies were taken and analysed as previously reported [10,11]. The 1997 TNM staging system  was used for clinical and pathological staging. Clinical staging consisted of a DRE and PSA measurement; abdominal and pelvic CT, and a bone scan, were obtained when the PSA level was > 10 ng/mL, the Gleason sum > 7, or possible clinical T3 disease was noted on a DRE. TAP and PFAP levels were measured by enzymatic assay (Sigma Diagnostic, St Louis, MO). The prostatic fraction was calculated as the tartrate-inhibitable fraction. ALP levels were measured by enzymatic assay. Serum levels of PSA were measured using the IMx assay (Abbott, Abbott Park, IL), with a lower limit of detection of 0.3 ng/mL. The first PSA value was obtained routinely 6 weeks after surgery; patients were then followed with a DRE and serum PSA levels every 6 months for 1 year, and then annually thereafter, although PSA was usually measured twice yearly. Cancer recurrence was defined as two consecutive serum PSA values of ≥0.3 ng/mL.
The data are presented as the number and percentage of patients, or median and interquartile range (IQR, 25–75th percentiles). Levels of TAP of > 5.4 IU/L, PFAP of > 1.2 IU/L and ALP of > 120 IU/L were classified as abnormally high. The relationship between abnormally high TAP, PFAP or ALP and measures of disease severity (pathological stage, Gleason score, perineural invasion, capsular penetration, positive margins, seminal vesicle involvement, and lymph node involvement) were assessed using Fisher's exact tests. The relationships with the PSA level before RRP was assessed using the Wilcoxon rank-sum test, and that with abnormally high TAP, PFAP or ALP and time to recurrence using log-rank tests.
- Top of page
- PATIENTS AND METHODS
Fifteen of the 195 patients who met the clinical criteria for the study cohort had neither AP nor ALP measured before RRP and thus 180 were available for analysis; in 164, 163, and 154 the TAP, PFAP and ALP were measured, respectively. The characteristics of the 180 patients are summarized in Table 1. The median duration of follow-up was 86 months (>7 years); 25 (14%) of the 180 patients developed recurrent cancer and nine died (one from prostate cancer and eight from other causes).
Table 1. The characteristics of 180 patients treated by RRP
|Characteristic||N (%) or median (IQR)|
|Age at diagnosis, years|| 64 (59–67)|
|PSA before RRP, ng/mL|| 7.0 (4.8–11.4)|
|TAP before RRP (164)|| 3.3 (2.6–4.0)|
|Abnormal (> 5.4 IU/L)|| 7 (4.3)|
|PFAP before RRP, IU/L (163)|| 0.7 (0.5–1.1)|
|Abnormal (> 1.2 IU/L)|| 33 (20)|
|ALP before RRP (n = 154)|| 81 (66–99)|
|Abnormal (> 120 IU/L)|| 13 (8)|
|T1b|| 3 (2)|
|T1c|| 43 (24)|
|T2a|| 91 (51)|
|T2b|| 43 (24)|
|Total biopsy cores obtained|| 6 (5–8)|
|Total biopsy cores positive|| 2 (1–3)|
|Percentage of biopsy cores positive|| 33 (20–57)|
|Biopsy total Gleason score|
|≤ 5|| 5 (3)|
|7|| 43 (24)|
|≥ 8|| 12 (7)|
|T2a|| 54 (30)|
|T2b|| 114 (63)|
|T3a|| 10 (6)|
|T3b|| 2 (1)|
|RRP total Gleason score|
|5|| 6 (3)|
|6|| 98 (54)|
|7|| 64 (36)|
|≥ 8|| 12 (7)|
|perineural invasion|| 61 (34)|
|capsular penetration|| 40 (22)|
|margins|| 48 (27)|
|seminal vesicle involvement|| 2 (1)|
|lymph node involvement|| 0 (0)|
The associations between abnormal TAP and PFAP and the severity of the disease, as indicated by the characteristics of the RRP specimen, are shown in Table 2. TAP was abnormal in seven patients (4%) and PFAP in 33 (20%), and neither were associated with any measure of disease severity, but the latter was associated with a higher PSA level before RRP (P = 0.02). ALP was abnormal in only 13 (8%) patients and not associated with any measure of disease severity. Neither abnormal TAP, PFAP nor ALP was associated with recurrence (P = 0.96, 0.45 and 0.41, respectively). However, for the three categories of PSA level (< 4, 4–10 and > 10 ng/mL) there was a significant difference in the likelihood of having recurrence (P < 0.001); Fig. 1 shows the univariate Kaplan-Meier curves for cancer recurrence according serum levels of PSA, TAP, PFAP and ALP.
Table 2. The severity of disease by abnormal TAP, PFAP and ALP
|Normal||Abnormal [P]||Normal||Abnormal [P]||Normal||Abnormal [P]|
|No. of patients||157|| 7 (4) [–]||130||33 (20) [–]||141||13 (8) [–]|
|age, years|| 64 (59–68)||64 (60–69) [–]|| 64 (60–68)||65 (59–67) [–]|| 63 (58–67)||65 (64–66) [–]|
|PSA before RRP, ng/mL|| 7 (5–12)|| 11 (5–12) [0.55]|| 7 (5–11)|| 9 (7–14) [0.02]|| 7 (5–11)|| 7 (7–11) [0.77]|
|Pathological stage|| || [0.81]|| || [0.57]|| || [0.16]|
|T2a|| 48 (31)|| 3 (43)|| 42 (32)|| 9 (27)|| 38 (27)|| 7 (54)|
|T2b|| 98 (62)|| 4 (57)|| 78 (60)||23 (70)|| 91 (65)|| 6 (46)|
|T3ab|| 11 (7)|| 0|| 10 (8)|| 1 (3)|| 12 (9)|| 0|
|Gleason score|| || [0.45]|| || [0.48]|| || [0.44]|
|5,6|| 89 (57)|| 3 (43)|| 73 (56)||18 (55)|| 81 (57)||10 (77)|
|7|| 57 (36)|| 3 (43)|| 49 (38)|| 11 (33)|| 50 (35)|| 3 (23)|
|8,9|| 11 (7)|| 1 (14)|| 8 (6)|| 4 (12)|| 10 (7)|| 0 (0)|
|perineural invasion|| 49 (31)|| 2 (29) [1.0]|| 41 (32)||10 (30) [1.0]|| 47 (33)|| 5 (38) [0.76]|
|capsular penetration|| 38 (24)|| 1 (14) [1.0]|| 31 (24)|| 8 (24) [1.0]|| 29 (21)|| 5 (38) [0.16]|
|margins|| 42 (27)|| 0 [0.19]|| 33 (25)|| 9 (27) [0.83]|| 41 (29)|| 4 (31) [1.0]|
|seminal vesicle involvement|| 2 (1)|| 0 [1.0]|| 2 (2)|| 0 [1.0]|| 2 (1)|| 0 [1.0]|
|lymph node involvement|| 0|| 0 || 0|| 0 || 0|| 0|
Figure 1. Kaplan-Meier estimates of the time to disease recurrence after RRP according to preoperative serum levels; P was determined by a log-rank test. (A) PSA, ng/mL (green < 4, red 4–10, light green > 10) P < 0.0001; (B) TAP, IU/L (green, < 5.4, red > 5.4) P = 0.96; (C) PFAP, IU/L (red, < 1.2, green > 1.2) P = 0.45; (D) ALP, IU/L (green, < 120, red £ 120) P = 0.41.
Download figure to PowerPoint
- Top of page
- PATIENTS AND METHODS
The role of preoperative levels of AP in determining the likelihood of extraprostatic disease and recurrence in patients who have undergone RRP remains controversial. For many years enzymatic serum AP was the only marker of prostate cancer and its levels were directly correlated with the extension of the disease, i.e. extraprostatic extension and metastatic disease. However, since Stamey et al. reported the clinical utility of PSA in detecting prostate cancer and its association with extension of the disease, AP has lost its utility as a tumour marker and currently is used infrequently in evaluating patients with prostate cancer.
Throughout the long history of the use of AP in urological practice there have been many studies that have shown an association between AP levels and the likelihood of extraprostatic extension. In 1984, Whitesel et al. found that patients with high preoperative enzymatic PAP and clinically localized prostate cancer had a higher likelihood of developing lymph node and bone metastases.
Bahnson and Catalona , in a retrospective study, found that high enzymatic serum AP was associated with an increased likelihood of tumour extension beyond the prostate. Oesterling et al. showed that enzymatic AP could predict capsular penetration, but not seminal vesicle and lymph node involvement. Salo and Rannikko  found that the elevation of PAP predicted both local extension outside the prostate and lymph node metastases, suggesting that radioimmunoassay was more sensitive in detecting extraprostatic spread than the enzymatic method. More recently, Moul et al. using a radioimmunoassay, and Han et al. using an enzymatic assay, reported a role for PAP and serum AP in predicting prostate cancer recurrence after RRP. Despite such studies indicating a role for AP in prostate cancer staging and outcome, some authors have suggested that the use of preoperative AP is unnecessary. For example, Stamey et al. compared the clinical usefulness of PSA and radioimmunoassayed PAP in patients with clinically localized prostate cancer. They concluded that PSA was more sensitive than PAP for detecting prostate cancer, was more useful in detecting the residual and early recurrence of tumours, and in monitoring the responses to radiation therapy. Likewise Burnett et al. found that enzymatic serum AP provided unique information in only 0.9% of their study cohort; the remaining patients could be staged by DRE and PSA levels, and therefore they concluded that serum AP was no longer mandatory before RRP. However, they did not evaluate cancer recurrence after RRP with serum AP levels.
To address this controversy, the present data, which were obtained after the introduction of PSA testing, showed no association between enzymatic TAP (or PFAP) and the pathological stage and/or outcome of patients after RRP, while simultaneously finding that PSA levels are clearly associated with the outcomes of the same patients (Fig. 1). Interestingly, high PFAP was associated with higher levels of preoperative PSA (P = 0.02) but abnormal PFAP was not associated with pathological stage or grade (Table 2). Therefore, from the present results, enzymatic TAP and PFAP are not useful in predicting disease severity or recurrence.
The role of ALP as a serum marker in patients with bone and liver metastasis is well known and it has been shown that BAP (which in healthy individuals contributes to < 40% of ALP ) has a complementary role to that of PSA in the diagnosis of bone metastases in patients with advanced prostate cancer . Morote et al. reported that the negative predictive value of a PSA level of < 20 ng/mL and BAP of < 20 ng/mL was 100% in detecting the lack of bone metastases in a series of 140 patients with prostate cancer. Murphy et al. showed that ALP and BAP had a significant discriminating ability for positive and negative bone scans. Lorente et al. also reported the clinical utility of BAP in avoiding unnecessary bone scans. Similar to the study by Wolff et al., < 10% of patients in the present cohort had abnormal ALP levels.
In another study, Wymenga et al. reported the clinical utility of ALP in predicting positive bone scans. These studies show the clinical utility of ALP and BAP levels in predicting which patients have possible bone metastases, but ALP and BAP have no role in evaluating patients with clinically localized prostate cancer. We found no association between ALP levels and PSA disease-free recurrence after RRP (Fig. 1D).
The main and unique conclusion of the present study is that enzymatic TAP, PFAP and ALP have no role in predicting the severity of disease and PSA disease-free recurrence after RRP in patients with clinically localized carcinoma in the present era of PSA testing. PSA levels before RRP can be used to differentiate the outcome for the same group of patients (Fig. 1A). These findings could be explained by stage migration of clinically detected prostate cancer caused by the use of PSA to detect prostate cancer [22,23]. TAP, PFAP and ALP may be more useful in study cohorts where prostate cancer is locally advanced, e.g. Moul et al. found that PAP was a predictor of recurrence after RRP in a surgical series in which 55% of the patients had extraprostatic disease (pT3–pT4). Likewise Han et al. reported a role for enzymatic AP in predicting recurrence after RRP in a series before PSA testing, in which 55% of the patients had extraprostatic disease after RRP.
In the present series (from 1990 to 1995) the median PSA was 7.0 ng/mL and < 10% of the patients had extraprostatic disease (Table 1); this is an absolute trend, because the median PSA for surgically treated patients with cancer has continued to decrease (e.g. the median PSA was 5.8 ng/mL in a series from 1990 to 2001).
Finally, we did not use multivariate analysis for cancer recurrence because the univariate analysis showed no relationship between the levels of enzymatic TAP, PFAP and ALP with pathological stage and likelihood of recurrence after RRP.
In conclusion, preoperative serum levels of enzymatic TAP, PFAP and ALP are not predictors of severity of disease and cancer recurrence after RRP for patients with clinically localized prostate cancer. PSA is generally better for predicting the therapeutic response and therefore the use of preoperative TAP, PFAP and ALP is not recommended routinely for patients with clinically localized prostate cancer.